BIO Web of Conferences 3, 02003 (2014)
DOI: 10.1051/bioconf/20140302003
c Owned by the authors, published by EDP Sciences, 2014
Indigenous Saccharomyces cerevisiae yeasts as a source of biodiversity
for the selection of starters for specific fermentations
Angela Capece, Rossana Romaniello, Rocchina Pietrafesa, and Patrizia Romano
Scuola di Scienze Agrarie, Forestali, Alimentari ed Ambientali, Università degli Studi della Basilicata, 85100 Potenza,
Italy
Abstract. The long-time studies on wine yeasts have determined a wide diffusion of inoculated fermentations
by commercial starters, mainly of Saccharomyces. Although the use of starter cultures has improved the
reproducibility of wine quality, the main drawback to this practice is the lack of the typical traits of wines
produced by spontaneous fermentation. These findings have stimulated wine-researchers and wine-makers
towards the selection of autochthonous strains as starter cultures. The objective of this study was to investigate the
biodiversity of 167 S. cerevisiae yeasts, isolated from spontaneous fermentation of grapes. The genetic variability
of isolates was evaluated by PCR amplification of inter-δ region with primer pair δ2/δ12. The same isolates
were investigated for characteristics of oenological interest, such as resistance to sulphur dioxide, ethanol and
copper and hydrogen sulphide production. On the basis of technological and molecular results, 20 strains were
chosen and tested into inoculated fermentations at laboratory scale. The experimental wines were analyzed for the
content of some by-products correlated to wine aroma, such as higher alcohols, acetaldehyde, ethyl acetate and
acetic acid. One selected strain was used as starter culture to perform fermentation at cellar level. The selection
program followed during this research project represents an optimal combination between two different trends
in modern winemaking: the use of S. cerevisiae as starter cultures and the starter culture selection for specific
fermentations.
1. Introduction
The conversion of grape must into wine is promoted by a
fermentation process naturally carried out by indigenous
yeasts [1]. Until about the 1980s, the contribution of
yeasts to wine production was seen as a relatively
simplistic concept. The main function of wine yeasts is
to guarantee the rapid and complete conversion of grape
sugar into ethanol, carbon dioxide, and many secondary
metabolites, avoiding the production of off-flavours.
Spontaneous fermentation is characterized by the activity
of different yeast species/strains, even if almost invariably
it is dominated by strains of the yeast Saccharomyces
cerevisiae. It is well known that the diversity of native
yeast strains is responsible for the production of wines
with different qualities and peculiar flavours. Yeast species
and, within each species, different strains exhibit wide
differences in volatile compound production, accounting
for the differences in composition and in taste of wine.
Although this yeast diversity can contribute to the wine
complexity and can produce unique-flavoured wines [2,
3], the dynamics of a spontaneous fermentation is often
unpredictable and some non-Saccharomyces species can
also produce undesirable compounds. Therefore, in the
wine-making industry the growth of undesirable yeasts is
controlled by addition of sulphur dioxide to musts and
inoculation with selected strains of Saccharomyces, mainly
S. cerevisiae.
The S. cerevisiae strains, involved in fermentation,
play an important role in the characteristics of wine [4, 5].
Although many flavour components derive directly from
the grapes, the essential part of a wine flavour is produced
during the alcoholic fermentation. In fact, the composition
and the sensory quality of the resulting wine are due to
the diversity of S. cerevisiae strains. Different strains of
S. cerevisiae can produce significantly different flavour
profiles when fermenting the same must [6] and this is
a consequence of the differential ability of wine yeast
strains to release varietal volatile compounds from grape
precursors and the strain-specific capability to de novo
synthesise yeast-derived volatile compounds [7].
Today, most wine is produced using selected commercial strains of Saccharomyces, but many wine-researchers
and winemakers prefer the use of selected autochthonous
strains of S. cerevisiae as starters. In fact, it seems that the
use of commercial dried yeasts reduced the biodiversity
of strains performing natural fermentation, and, as a
consequence of this, a reduction of the resulting wine
complexity. Actually, also small wineries are interested
in the selection of yeasts from their own environment for
use as starter cultures. Extensive ecological surveys using
molecular methods have been carried out with the aim of
selecting new yeasts better adapted to local fermentation
conditions [8]. The indigenous S. cerevisiae strains are
better acclimated to micro-area conditions of the wine
production region and therefore they can more easily
This is an Open Access article distributed under the terms of the Creative Commons Attribution License 4.0, which permits unrestricted use, distribution,
and reproduction in any medium, provided the original work is properly cited.
Article available at http://www.bio-conferences.org or http://dx.doi.org/10.1051/bioconf/20140302003
BIO Web of Conferences
dominate on the natural biota. Furthermore, the use of
locally selected yeast strains with strain-specific metabolic
characteristics could positively affect the final quality of
wine [9, 10] and ensure the maintenance of the typical
sensory properties of wines deriving from a specific area.
The selection protocol of indigenous strains, which
have to be used as starter cultures for specific
fermentations, includes the study of different parameters.
Among the desirable and traditional oenological criteria,
the following are the most frequently evaluated during
the selection protocol: high fermentation performance,
resistance to, and low production of, sulphur dioxide; low
production of hydrogen sulphide and low volatile acidity;
resistance to ethanol; positive influence on wine aroma.
In recent studies addressed to the selection of indigenous
starter cultures, another trait considered was the evaluation
of strain ability to dominate natural yeast population (strain
implantation), always present during must fermentation
[11, 12]. “Matera DOC” is a little wine-production of
Basilicata, Southern Italy, which awarded its Registered
Designation of Origin (DOC) classification in 2005. The
types of wines produced are six, three red and three
white. Some producers carry out spontaneous fermentation
of grape musts, whereas other cellars utilize commercial
yeasts strains as starters to reduce the risk of wine spoilage
and uniform wine quality.
The aim of this study was to investigate the biodiversity
and the oenological properties of S. cerevisiae isolated
during spontaneous fermentation of grapes collected in
this specific area in order to select those strains with
specific characteristics and well adapted to the cellar
environment to be used as starters in winemaking. The
selection program followed during this research project
represents an optimal combination between two different
trends in modern winemaking: the use of S. cerevisiae
starter cultures which can produce wine characterized by a
reproducible quality and the selection of the starter culture
for specific fermentations in function of the vine variety
characteristics.
2. Material and methods
2.1. Yeasts
One hundred and sixty-seven Saccharomyces cerevisiae
isolates were studied. The yeasts were previously isolated
from spontaneously fermented grapes (Aglianico variety),
collected from vineyard grown under organic farming
methods, cultivated near Pollino National Park (Basilicata
Region, Southern Italy). The isolates were maintained on
YPD medium (1% (w/v) yeast extract, 2% (w/v) peptone,
2% (w/v) glucose, 2% (w/v) agar).
2.2. Technological characterization
The isolates were tested for their tolerance to different
antimicrobial compounds, such as ethanol (EtOH), sulphur
dioxide (SO2 ) and copper sulphate (CuSO4 ). These tests
were performed as described by Mauriello et al. [6].
Hydrogen sulphide (H2 S) production was evaluated by
inoculating the yeasts on bismuth-containing indicator
medium B I GG Y agar (Oxoid) and the plates were
incubated at 26 ◦ C for 2 days. On this medium the
production level of H2 S is related to browning of yeast
colonies: H2 S-positive strains exhibit brown or black
colonies, while H2 S-negative colonies are white.
All the data obtained by technological characterization
were converted into non-dimensional values, assigning for
each isolate the following values:
- 0 for parameters exhibited at low level,
- 1 at middle level,
- 2 at high level. The values were submitted to the
cluster analysis, using the Paired Group method with
Euclidean distance. The statistical package used was
PAST software ver. 1.90 [13].
2.3. Genotypic characterization
The genetic variability among the 167 isolates was
evaluated by amplification of inter-δ region with primer
pair δ2-δ12 [14, 15], following the protocol described by
Capece et al. [10].
2.4. Fermentation trials at laboratory-scale with
selected S. cerevisiae strains
Twenty selected S. cerevisiae isolates were tested in
inoculated fermentations at laboratory scale in comparison
to S. cerevisiae commercial strain used in the cellars.
The fermentations were performed in 130-ml Erlenmeyer
flasks filled with 100 ml of grape must, added with 50 mg
l−1 of SO2 . Each strain was inoculated in grape must at a
concentration of 106 cells ml-1, from a pre-culture grown
for 48 h in the same must. The fermentation was performed
at 26 ◦ C and the fermentative course was monitored by
measuring weight loss, determined by carbon dioxide
evolution during the process. At the end of the process
(stable weight), the wine samples were refrigerated at 4 ◦ C
to clarify the wine, racked and stored at −20 ◦ C until
required for analysis. All the experiments were performed
in duplicate.
During this experiment, the qualitative amount of H2 S
formed during fermentation was determined by evaluating
the browning degree of commercial lead acetate test strips,
inserted in the top of fermentation flasks. The degree of
blackening of the strips correlates to the amount of H2 S
produced during fermentation [16].
The content of some secondary compounds influencing
wine aroma were analyzed by direct injection gas chromatography of 1 µl of experimental wines, by following
the method described in Capece et al. [12]. Levels of these
compounds were quantified by internal standardization
(calibration curves) using Agilent ChemStation Software.
Levels of secondary compounds determined in the
experimental wines were submitted to statistical analysis
by descriptive Box plots and whiskers, using PAST
software ver. 1.90. The experimental wines were analyzed
for the total and free SO2 content, which was measured
iodo-metrically by the Ripper procedure [17].
2.5. Inoculated fermentations at pilot scale
Fermentations at pilot scale were carried out in a cellar
producing “DOC Matera wine” during the 2013 vintage.
02003-p.2
37th World Congress of Vine and Wine and 12th General Assembly of the OIV
One indigenous strain (selected during this study) and
the commercial strain commonly used in the cellar were
tested. The fermentations were performed in sulphite
(50 mg l−1 ) Aglianico grape must (240 g l−1 sugar,
pH 3.5,), by inoculating 106 cells ml−1 in 100−1 -vats
and the fermentation processes were monitored daily
by determining sugar and temperature. The implantation
ability of inoculated strains during the fermentative
process was tested by following the protocol described
by Capece et al. [10]. Conventional chemical parameters
of wines, such as total acidity, volatile acidity, glucose
and fructose, alcohol content, lactic and malic acids,
were measured using Fourier Transfer Infrared WineScan
(FOSS, Hillerød, Denmark), whereas the content of
secondary compounds affecting wine aroma was detected
as previously reported.
3. Results
3.1. Technological characterization
In the first step of this research, 167 S. cerevisiae
yeasts, isolated from spontaneous fermentation of Aglianico grapes, were investigated for characteristics of
oenological interest, such as resistance to chemical
compounds potentially present during the process (SO2 ,
EtOH, CuSO4 ) and H2 S production. All the data
related to the technological characterization were statistically elaborated, obtaining the dendrogram reported in
Fig. 1. As shown in the figure, the strains were distributed
in 14 groups, some of which were composed by one/two
isolates (1, 3, 5, 7, 12, 13 and 14). The most numerous
groups were 4, 7 and 9, which grouped 31, 30 and 43
isolates, respectively (Table 1). Generally, the isolates
exhibited a high tolerance to SO2 (Table 1); in fact, the
majority tolerated the highest dose tested (300 mg l−1 ),
whereas only 3 isolates were inhibited by concentrations
higher than 100 mg l−1 (groups 1 and 7, Fig. 1). On the
contrary, the main percentage of isolates was characterized
by low EtOH tolerance (growth in presence of 12% v/v)
(all the groups reported in Fig. 1, except 8, 9 and 13)
and only few isolates (belonging to group 10) tolerated
the highest dose tested (18% v/v). A certain variability
was found for copper resistance; about 50% of isolates
were inhibited by lowest doses tested (0–100 mmol l−1 )
and the tolerance levels of other isolates were distributed
among the other concentrations tested. As regards the H2 S
production, generally the isolates were characterized by
medium production level as, after growth on B I GG Y agar,
they developed colonies characterized by brown color.
Only the strain M1-47 (group 5, Fig. 1) was no-H2 S
producer; in fact, this strain developed white colonies on
B I GG Y agar.
3.2. Genotypic characterization
The evaluation of genetic polymorphisms among 167
isolates was carried out by PCR analysis of inter-δregion
by using the primer pair δ2/δ12. Six different interdeltaprofiles were found among the 167 yeasts; the profiles
obtained are reported in Fig. 2, in which a different letter
was assigned to each pattern that differed from the others
in at least one intense band.
Figure 1. Dendrogram obtained after a hierarchical agglomerative cluster analysis performed on data of technological
characterization of 167 S. cerevisiae isolates.
The profile “a” was the most common (exhibited
by 59% of analyzed isolates), followed by pattern “b”
(29% of yeasts), whereas other profiles were exhibited
by few strains. The profiles “d” and “e” were specific of
single strains. The comparison of results obtained with
technological and genetic characterization revealed that
isolates grouped in the same technological group showed
the same molecular profile (see Table 1), although isolates
sharing the same molecular profile were subdivided
in different technological groups. For example, isolates
exhibiting the inter-δ profile “a” were distributed in 6
02003-p.3
BIO Web of Conferences
Table 1. Characteristics of 167 Saccharomyces cerevisiae isolates, grouped according to technological characterization.
Groupa
1
2
3
4
5
6
7
8
9
10
11
12
13
14
EtOH*
12
12
12
12
12
12
12
14
14–16
18
12
12
14
12
Resistance
SO2 **
Cu***
100
200
200
200
300
100
300
200
300
200
200
< 100
100
100
200
0–100
300
100
300
100
300
400–500
300
200
300
400
300
400
H2 Sb
M
M
M
M
L
M
M
M
M
M
H
H
M
M
N◦ of
Isolates
1
15
2
31
1
30
2
7
43
17
14
1
1
2
Inter-δ
profile
a
b
a1
b
b
a
a
a
a
a
c
d
e
b
a = Technological groups, reported in Fig. 1; b = production evaluated on BIGGY agar; M = medium; L = low; H = high; * (% v/v); ** (mg/L); *** (µmol/L of CuSO4 ).
Figure 2. Molecular profiles obtained by amplification of interδ region with primer pair δ2/δ12 of indigenous S. cerevisiae
isolates. M: 100 bp marker (Biolabs).
technological groups (1, 6, 7, 8, 9 and 10, Table 1).
Probably, the analysis of the inter-δ region by using only
a primer pair was not enough to discriminate the isolates,
as previously reported [10, 18]. At this purpose, the use of
other typing molecular techniques, which analyze different
regions of yeast genetic patrimony, will be necessary to
correctly evaluate the genetic variability among the 167 S.
cerevisiae isolates.
3.2.1. Fermentation trials at laboratory-scale with
selected S. cerevisiae strains
On the basis of the technological characteristics, twenty
S. cerevisiae isolates, representative of the different
clusters reported in Fig. 1, were selected and subjected
to further characterization. In order to analyze the
fermentative performance of selected strains, the S.
cerevisiae isolates were tested in inoculated fermentations
at laboratory scale in comparison to a commercial starter
widely used in the wineries (CS). The fermentations
were performed in the same grape must from which the
yeasts were isolated; the process course was monitored by
evaluating carbon dioxide evolution and the process was
completed after 12–14 days. The Figs. 3a and b reports for
each strain the fermentative rate, expressed as amount of
Figure 3. Fermentation rate expressed as g CO2 /day, measured at
the first 2 days (a) and at the end (b) of the fermentation.
CO2 produced for each fermentation day, measured at the
first two days of fermentation (Fig. 3a) and at the end of
the process (Fig. 3b). High variability in fermentation rate
was recorded during the first step of the process. Some
strains (M1-109, M3-1 and M3-66) produced very low
level of CO2 /day, whereas other strains (M1-101, M1-57
and M3-59) produced more than 1 g CO2 /day and these
values were higher than the fermentation rate exhibited
by the commercial starter. When the fermentation rate
was measured at the end of the process (Fig. 3b), the
differences among the strains were reduced and indigenous
02003-p.4
37th World Congress of Vine and Wine and 12th General Assembly of the OIV
Table 2. Characteristics of experimental wines obtained by inoculating 20 indigenous S. cerevisiae strains in comparison to commercial
starter (CS). ADE = acetaldehyde; N-PR= n-propanol; ACET = acetic acid; ISM = isoamyl alcohol.
Strains
M1-22
M1-27
M1-38
M1-47
M1-55
M1-57
M1-82
M1-101
M1-104
M1-107
M1-109
M1-110
M3-1
M3-37
M3-54
M3-59
M3-60
M3-66
M3-71
M3-80
CS
Sulphur compounds*
SO2 (T) SO2 (F) H2 S
12
12
2
12,8
9,6
2
10,4
9,6
3
55,2
9,6
0
15,2
10,4
0
12
12
1
12
12
1
12
11,2
4
10,4
9,6
4
9,6
9,6
4
12,8
11,2
1
11,2
11,2
1
14,4
10,4
2
16,8
10,4
1
14,4
12,8
3
14,4
12
2
12
10,4
0
14,4
12,8
1
13,6
11,2
2
16,8
12
2
13,6
12
3
Aromatic compounds (mg l−1 )
ADE
N-PR
ACET
21,38
29,86
570,28
72,54
33,07
206,52
48,74
27,15
143,51
136,41
85,65
217,48
44,46
44,37
180,00
23,31
33,91
299,01
24,27
29,32
414,81
82,05
25,42
165,18
89,77
34,64
228,89
60,68
28,88
235,51
31,33
32,94
243,23
44,15
32,05
373,02
36,15
36,40
224,58
43,79
40,67
166,16
42,44
43,04
241,89
38,47
34,03
227,72
21,79
46,35
239,40
30,25
38,30
210,42
52,67
42,92
210,25
47,12
42,68
232,23
39,61
34,31
359,89
ISM
241,15
185,38
189,66
153,02
190,65
199,12
238,72
163,95
169,06
176,05
194,54
243,64
179,92
179,03
180,27
208,91
146,22
198,99
170,76
191,46
207,66
* Concentration of total (T) and free SO2 (F) was expressed as mg l−1 ; H2 S production was expressed as arbitrary values in function of blackening degree of lead acetate strips.
Figure 4. Box plot representing the variability of secondary
compounds determined in the experimental wines produced
by inoculating 20 S. cerevisiae strains. ADE = acetaldehyde;
ETAC = ethyl acetate; N-PR = n-propanol; ISB = isobutanol;
ACET = acetic acid; D-AM = D-amyl alcohol; ISM = isoamyl
alcohol.
strains exhibiting fermentation rate comparable to those of
commercial starter were found (i.e. M1-22, M3-59, M1-57
and M1-47). The lowest value was exhibited by the same
strains showing the lowest fermentation rate at the first two
days of the process.
The experimental wines obtained were analyzed
for the content of some by-products correlated to the
organoleptic quality of wine. The metabolites determined
by gas-chromatographic analysis were: higher alcohols (npropanol, isobutanol, amyl alcohols), acetaldehyde, ethyl
acetate and acetic acid. The determination of metabolite
levels showed considerable differences in the obtained
wines (Fig. 4); in particular, high variability in the levels of
acetic acid, isoamyl alcohol and acetaldehyde was found.
Although the variability found in the production level
of acetic acid and isoamyl alcohol, in all the obtained
wines the level of these compounds was comprised in the
desirable range. In fact, the maximum amount of acetic
acid detected in the experimental wines was 570 mg l−1 ,
below the level considered negative (> 600 mg l−1 ) and
the maximum amount of isoamyl alcohol was 240 mg l−1
and this compound exert a negative influence on wine
aroma when the level exceeds 300 mg l−1 . As regards
the acetaldehyde content, only the wine produced by
inoculating the strain M1-47 contained a very high amount
of this compound (135 mg l−1 ).
As regards the SO2 content of wine (Table 2), the
maximum amount detected was 16.8 mg l−1 of total SO2 ;
in the majority of wines, the SO2 was present in free form.
A different behaviour was found for the strain M147; in fact, the wine produced by inoculating this
strain contained the highest level of SO2 (55.2 mg l−1 );
furthermore, in this wine the majority of SO2 is present
in bound form as the level of free SO2 is very low
(9,6 mg l−1 ). This might be due to the high production
level of acetaldehyde (Table 2), which represents one of
the principal sulphite-binding compounds. Furthermore,
this strain is characterized by non-production of hydrogen
sulphide (Table 1, group 5) and the highest production
level of n-propanol.These results confirmed the data
previously obtained [16], reporting the close relationship
between high production of n-propanol and strain
incapacity to produce hydrogen sulphide.
The results regarding the evaluation during fermentation of H2 S production by lead acetate strips were reported
in Table 2, in which arbitrary values were assigned in
function of the degree of blackening of lead acetate strips.
The value ranged from 0, assigned to the strains producing
very small amount of H2 S with the strips remaining white
02003-p.5
BIO Web of Conferences
Table 3. Conventional parameters and main secondary compounds concentrations of wines elaborated at cellar scale
with indigenous (M3-59) and commercial starter (CS). The
concentration was expressed as: a = g l−1 , b = %v/v; c = mg l−1 .
Parameters
Total aciditya
Volatile aciditya
Glucose + fructosea
Ethanolb
Malic Acida
Lactic acida
Acetaldehydec
Ethyl acetatec
N-propanolc
Isobutanolc
D-amyl alcoholc
Isoamyl alcoholc
M3-59
7,9
0,40
0,0
13,81
2,37
0,33
16,93
24,30
28,75
30,57
97,32
145,58
CS
6,6
0,42
0,0
13,44
1,17
1,05
18,25
26,82
27,78
59,88
165,19
322,60
were characterized by fermentative performance similar
or superior to commercial starter utilized by the wine
cellar. The fermentation at pilot scale confirmed that the
selected indigenous strain possesses oenological properties
comparable with commercial starter; furthermore, the wine
produced by the indigenous starter was characterized
by a more balanced aroma than the wine produced by
commercial starter.
Further experiments are in progress in order to
test other indigenous strains as starter culture in real
winemaking conditions. The strain M1-47 seems to be of
interest for winemaking, being non H2 S producing, but
further studies are necessary in order to test its behaviour
at cellar level.
This work was supported by project PIF-LIELUC (Misura 124
PIF Vini di Lucania, PSR Basilicata 2007–2013 “Lieviti Indigeni
per Vini Lucani” N. 94752044753).
or near white, to 4, assigned to the strains producing
very high amount of H2 S with strips very black. Different
behaviours were found, with strains producing very small
amounts of H2 S, such as M1-47, M1-55 and M3-60, and
strains producing a considerable amount of this sulphur
compound, such as M1-101, M1-104 and M1-107.
3.3. Inoculated fermentations at pilot scale
On the basis of these results, the strain M3-59 was selected
for fermentation at pilot scale as this indigenous strain possessed the following suitable oenological characteristics:
high resistance to antimicrobial compounds and medium
H2 S production (technological group 13, Table 1); high
fermentation rate (Fig. 3) and balanced production of
secondary compounds (Table 2). This strain was tested in
comparison to the commercial strain, commonly used in
the cellar. Both the starters displayed ability to exhaust
must sugars and to dominate the fermentation, with a
strain implantation of 100% in both the fermentations. The
other conventional parameters were very similar in the
two wines and were comprised in the acceptable levels.
High differences between the two wines were found for
the content of amyl alcohols, which was very high in the
wine obtained by inoculating the commercial starter. In
particular, the content of isoamyl alcohol in the wine by
commercial starter was 322 mg l−1 , a value slightly higher
than 300 mg l−1 , the level considered negative for wine
aroma.
4. Conclusions
This study was focused on the biodiversity of S. cerevisiae
yeasts isolated from spontaneously fermenting grapes,
collected from a specific production area of “DOC Matera”
wine. It has to be underlined that studies on yeasts isolated
from this environment had not been performed before.
The characterization for traits of oenological interest
revealed that some of these strains possess interesting
technological traits and could represent starter cultures
available for winemakers, who are addressed to production
of quality premium wines maintaining the differential
properties of their own area. Different indigenous strains
References
[1] G. H. Fleet, J. Food Microbiol. 86, 11 (2003)
[2] M. Combina, A. Elia, L. Mercado, C. Catania, A.
Ganga, C. Martinez, Int. J. Food Microbiol. 99, 237
(2005)
[3] K. Zott, C. Miot-Sertier, O. Claisse, A. LonvaudFunel, I. Masneuf-Pomarede, Int .J. Food Microbiol.
125, 197 (2008)
[4] E. S. King, J. H. Swiegers, B. Travis, I. L. Francis,
S. E. P. Bastian, I. S. Petrorius, J. Agric. Food Chem.
56, 10829 (2008)
[5] M. G. Lambrechts, I. S. Pretorius, S.Afr. J. Enol.
Vitic. 21, 97 (2000)
[6] G. Mauriello, A. Capece, M. D’Auria, T. GardeCerdán, P. Romano, Food Microbiol. 26, 246 (2009)
[7] M. Vilanova, C. Sieiro, J. Ind. Microbiol. Biotechnol.
33, 929 (2006)
[8] H. Csoma, N. Zakany, A. Capece, P. Romano, M.
Sipiczki, Int. J. Food Microbiol. 140, 239 (2010)
[9] A. Capece, R. Romaniello, G. Siesto, R. Pietrafesa,
C. Massari, C. Poeta, P. Romano, Int. J. Food
Microbiol. 144, 187 (2010)
[10] A. Capece, R. Romaniello, G. Siesto, P. Romano
Food Microbiol. 31, 159 (2012)
[11] N. Barrajón, M. Arévalo-Villena, L. J. Rodrı́guezAragón, A. Briones, Food Control, 20, 778 (2009)
[12] A. Capece, G. Siesto, R. Romaniello, V. M. Lagreca,
R. Pietrafesa, A. Calabretti, P. Romano P., LWT Food Sci. Technol. 54, 485 (2013)
[13] Ø.
Hammer,
D.
Harper,
P.D.
Ryan
http://folk.uio.no/ohammer/past. (2001)
[14] F. Ness, F. Lavallée, D. Dubourdieu, M. Aigle, L.
Dulau, J. Sci. Food Agric. 62, 89 (1993)
[15] J. Legras, F. Karst, FEMS Microbiol. Lett. 221, 249
(2003)
[16] P. Giudici, R.E. Kunkee, Am. J. Enol. Viticult. 45,
107 (1994)
[17] M.A. Amerine, C.S. Ough, Methods for Analysis
of Musts and Wine (New York: Wiley-Interscience
Publication, 1974)
[18] G. Siesto, A. Capece, M. Sipiczki, H. Csoma,
P. Romano P., Ann. Microbiol. 63, 661 (2013)
02003-p.6